The conventional secret key backup method is implemented by using the secret distribution technique.
This is the technique being able to implement safe backup while maintaining the contents of the secret key confidential by dividing the secret key into plural fragments of secret key, and storing each of the plural fragments of secret key by plural administrators (for example, Non-patent document 1).
This method is characterized in that, in addition to specifying the number of divisions N into which the secret key is divided, namely, the number of administrators, it is possible to specify the number of restoration enabling pieces k which shows how many number of fragments of secret key (the administrators) required for restoring the secret key out of the number; and for example, by specifying the numbers as k<N, it is possible to restore the secret key even if one administrator loses a fragment of the secret key.
Further, another secret distribution technique has been proposed to enable to restore the secret key with an arbitrary combination of administrators by devising a delivering method of divided fragments of secret key to the administrators (for example, Patent document 1).
In this method, when there are N administrators P1 through PN, an access structure, which specifies administrators whose fragments of secret key can collectively restore the secret key, is specified.
For example, in order to show that when there are three administrators (P1, P2, P3), the restoration of secret key requires fragments of secret key of two administrators, and one of which is the administrator P1, the access structure is specified as {{P1, P2}, {P1, P3}}.
Then, the secret key is divided into N fragments of secret key by implementing secret distribution of the secret key based on the access structure.
Then, by distributing fragments of secret key to respective administrators, it is possible to restore the secret key only when the administrators of a combination specified by the access structure get together.
On the other hand, Role-Based Access Control (RBAC) has been used in the intra-firm information system of recent years (for example, Non-patent document 2).
This is a mechanism to define roles according to the role of the work or the organizational structure, and to specify the access authority for the role.
Moreover, by forming the role with hierarchical structure, it is possible to inherit the authority.
Further, by registering the user to the role, it is possible to carry out access control of the users.
If change in personnel occurs, it is sufficient to change only the registration of users for the role, so that the method is said to be suitable for the intra-firm information system.
For example, a general manager role and a manager role are created as roles; the authority to settle payment is assigned to the general manager role, the authority to prepare a disbursement voucher is assigned to the manager role. Then, a user of “General Manager Tanaka” is registered to the general manager role, and users of “Manager Suzuki” and “Manager Sato” are registered to the manager role. By this operation, it is possible to carry out access control so that Manager Suzuki or Manager Sato prepares a disbursement slip and General Manager Tanaka is authorized to settle the payment. Further, by making the general manager role inherit the authority of the manager role, it is also possible for General Manager Tanaka to prepare a disbursement. Further, when Takahashi newly becomes a manager, the access authority can be modified easily only by registering Manager Takahashi to the manager role.    Patent document 1: JP 2002-217891A, pp. 5-10    Non-patent document 1: A. Shamir, “How to Share a Secret”, Communications of the ACM, v. 22 n. 11, p. 612-613, November 1979    Non-patent document 2: David F. Ferraiolo, D. Richard Kuhn, and Ramaswamy Chandramouli, “Role-Based Access Control”, ARTECH HOUSE, INC., 2003, p. 6-16